It's June 23. The San Francisco Giants are 75 games into a dreadful season where they've won just 27 games. Mark Melancon, the closer who the Giants inked for 62 million dollars in the offseason, has appeared in just 31.7 percent of games in which he was on the active roster.
This number is currently the lowest percentage among all bullpen arms, whether they are still on the team or not.
- Steven Okert: 54.3%
- Hunter Strickland: 44.8%
- Josh Osich: 43.8%
- Derek Law: 43.1%
- George Kontos: 42.7%
- Sam Dyson: 41.7%
- Bryan Morris: 40.0%
- Cory Gearrin: 37.3%
- Neil Ramirez: 36.0%
- Mark Melancon: 31.7%
You might say, "Well, the Giants just haven't had enough save opportunities!"
Because he is the closer, right? He comes into the game to get saves. That's his job.
But he's still the best reliever on the staff. Forget about his 2017 numbers. He owned a 1.64 ERA, 0.89 WHIP, and reasonable 24.1 K% in 2016 split between Pittsburgh and Washington. His highest ERA in any season since 2013 was 2.23 in 2015. And during all four seasons between 2013 and 2016, his WHIP sat below 0.96.
Thrown in with the names above, he's without question the best of the bunch. So, doesn't that mean he should pitch in the most games? Or at the absolute very least be in the top three in percentage of games used?
He's dead last. 62 million over four years pitching in 4.3% fewer games than the guy in second to last and in 8.3% fewer games than a guy signed to a minor-league contract to replace a guy previously signed to a minor-league contract.
Again, you'll say," The Giants are so bad that there haven't been any games to save!"
So it begs the question: Is this normal? Is this what is supposed to happen? Do team's best relievers who are also the closer get used less frequently if their team is bad and/or has fewer save opportunities?
Let's look at the 1997 San Diego Padres. They finished the season 76-86. Their closer and best reliever was Trevor Hoffman. And their manager? Bruce Bochy. How often was Hoffman used? Did the Padres have more save opportunities?
Fifty-five games revealed save situations in 1997 for the Padres, meaning 33.9% of the entire season resulted in a save chance for someone late in the game. Only 20 games have required saving in the Giants' 75 games, which is 26.7% of the season.
Does this justify using your closer less?
First we'll note Melancon's use on average days rest since 2013.
- 2013: 1.5
- 2014: 1.5
- 2015: 1.3
- 2016: 1.5
- 2017: 2.5
Keep that in mind as we progress.
Hoffman in 1997 appeared in 70 games. He was never not on the active roster, so that gives him a 43.2% mark in 162 games. The Padres were 5.5 games or more behind the division leader from April 26 until the end of the year. And it only moved below 7.0 games back five times. So, like the 2017 Giants, they were really never in contention from very early on.
Hoffman recorded 37 saves in 44 opportunities that year. So that leaves 26 more appearances that were not save situations, or 37.1% of his overall use. What are those appearances comprised of? Mop-up work to keep him fresh?
I'm going to sort all 26 non-save uses into separate categories.
- Home, 8th inning or later, tie game: 9
- Road, 8th inning or later, tied or behind by 1: 4
- 3+ days rest, behind 2+ runs: 4
- 0-2 days rest, ahead 4+ runs: 3
- 0-2 days rest, behind 2+ runs: 2
- Home, 9th inning or later, behind by 1: 2
- 3+ days rest, ahead 4+ runs: 2
Hoffman's average days rest in 1997 was 1.6. His average days rest a year later in 1998 when the Padres went 98-64 and lost in the World Series was 1.8. He, like 1997, did not hit the disabled list and was active for 162 games. He appeared in 66 games, which is a 40.7% mark.
Melancon has recorded 11 saves in 15 opportunities. So that leaves five more appearances that were not save situations, or 25.0% of his overall use. Now, employing those same categories, when has Melancon been used by Bochy in 2017 during non-save situations?
- Home, 8th inning or later, tie game: 3
- 3+ days rest, behind 2+ runs: 1
- 3+ days rest, ahead 4+ runs: 1
This leaves a 12.1% difference between the use of Hoffman in 1997 and the use of Melancon in 2017 — both in similarly losing seasons. To get Melancon up to 37.1%, on par with Hoffman in 1997, Bochy would have had to have used Melcanon four more times up until this point.
Remember Hoffman's days rest average of 1.6 versus Melancon's 2.5. Hoffman was used on four or more days of rest in only 13.0% of his outings. On the other hand, Melancon has been used on four or more days rest in 33.3% of his outings, including once on nine days rest.
Where might Bochy have been able to use Melancon this year that would be in line with his use of Hoffman in 1997?
- April 7 at San Diego: Down 7-6 in the eighth inning, Derek Law was used. Melancon had not pitched since April 2. Law pitched fine, but Melancon would now go into the next game on five days rest.
- April 15 vs. Colorado: Down 5-0 in the ninth inning, Ty Blach was used. Melancon had not pitched since April 10. Blach pitched fine, but Melancon would go into the next game on five days rest.
- April 23 at Colorado: Down 8-0 in the eighth inning, Hunter Strickland was used. Melancon had not pitched since April 16. Strickland pitched fine, but Melancon would go into the next game on five days rest.
- June 15 at Colorado: Hunter Strickland, having pitched the previous two nights, faced the Rockies' 4, 5, and 6 hitters in the bottom of the ninth inning in a tie game. Strickland gave up a single, got an out, and walked a guy. With runners on first and second and one out, Melancon, having not pitched since June 8, sat in the bullpen. The Giants lost two pitches later.
- June 17 at Colorado: Down 5-1 in the eighth inning on the road, Cory Gearrin was used. Melancon had still not pitched since June 8. Gearrin had pitched three previous times between Melancon's last use and June 17.
April 7 and June 15 fall into the "Road, 8th inning or later, tied or behind by 1" category, which Bochy utilized four times in 1997. He has yet to do it in 2017.
April 15, April 23, and June 17 fall into the "3+ days rest, behind 2+ runs" category, which happened four times in 1997, and has happened once in 2017.
June 17 is the most glaring lack of use. It was the perfect opportunity to get Melancon fresh for any potential future save situations, considering he would have been pitching on nine days rest if he pitched the next day.
He did pitch the next day — and he blew his fourth save of the year.
It appears, when just comparing 1997 to 2017, Bochy made a concerted effort with Hoffman to keep him fresh, using him in "mop-up" type situations. Six times Hoffman was used in either the "3+ days rest, behind 2+ runs" or "3+ days rest, ahead 4+ runs" category.
And during all four uses in the "Road, 8th inning or later, tied or behind by 1" category, Hoffman was pitching on three or more days rest, including once on six days rest.
These are perfect uses. Whether at home or on the road, pitching in the eighth inning or later when either tied or down one is a high leverage situation. When the closer hasn't pitched in three or more days, this is an ideal chance to keep him fresh while not having to throw him out there in mop-up work. You can essentially keep him fresh specifically for high-leverage work without it actually being a save opportunity.
Bochy has yet to use Melancon in this situation in 2017.
This isn't even addressing the subject of closer use before the eighth inning. All proposed uses of Melancon in 2017 are simply in an effort to keep him more fresh, and are in line with previous Bochy practices.
It's certainly possible that Bochy's philosophy on closer use has changed in the last 20 years. But you would think 20 years ago, he may have been even more traditional than the present time. It seems now he's more strictly held to reserve his closer only for save situations.